Ask Lemmy

I art. I do love ai for the lulz, however, actual commercial art? Absolutely not. It's not an end product. It's fun, it's inspiring.

Like anything art generators are a tool. One that can be very useful in a creative process, to convey an idea that is hard to present in text, to explore variations on a concept without having to draw something a hundred times, etc. It would be very difficult to argue that something like that has no valid uses.

However, as it stands the majority of the tools in place cost a fair bit of money to set up and run and so there is a high barrier to entry, and so the profits made from running them end up going primarily to those who are rich enough to set them up in the first place. Wealth inequality is a massive issue right now and so this sours a lot of people against these tools.

Many people also subjectively dislike AI art, which is a fair comment, as all art is subjective, but I don't think it necessarily helps anyone to debate over whether it looks good or not, that shouldn't be the issue here.

You could argue that the root of the problem is that most users of these tools will never consider the repercussions of paying for them, the people they are supporting are obscured behind many layers and it is impossible for the average consumer to know what the recipient will do with those funds.

Like any tool, these machines have created a new way for the already powerful to exploit the weak, it may be abstracted away behind closed doors but it is happening.

Not really, if they actually look good and doesn't have the uncanny valley stuff to it. But there should be rules on Lemmy (and hopefully other platforms too) to required images to be marked as AI.

I am fine with AI art as long as its properly credited to its creato. Not the person who wrote a prompt to generate the image, not the company that created the program. The AI should be credited in a way that no person could confuse it for something someone made

If thats too hard, banning AI art is also fine. I havent seen any real use for it

Nice try. I'm not helping you improve your art algorithms for free. You need to pay some art teachers for feedback like that.

I don't hate it, I think it has its uses, just like text generation. They're great for brainstorming ideas or quick unimportant stuff like RPG campaigns, so for example an in-game fake company logo or a poem to contain hints for the players.

However trying to use it for anything serious and final is stupid and dangerous. IMO any artist that had their art used to train a model should be able to claim royalties on anything created with that model, regardless of whether they can prove their art was used for the piece. And if the data used to train the model is not made public or can't be verified, then ANY artist can. Maybe just 1% of the profits direct or indirect of that art, so for example you used AI to generate part of an invitation for a party, 100 artists could start a lawsuit and take every single cent you earned from the party. After all you indirectly hired them, it's only fair they get paid, had you hired a single artist you could negotiate the price with them.

If I see a obviously AI generated picture as a thumbnail on youtube, I immediately block that creator. If I hear those awful AI voices reading text, same. If you want to share something with the world, put some effort into it.

Use case seems to just be dicking around, and that is just not worth the resources we pour into it.

I like playing around with it myself but I never upload it I just keep it on my computer cuz it's neat so I don't get why anyone else would upload AI generated stuff online

If it came from stealing actual artists' work then I hate it. If they somehow generated it using all fairly sourced data then I don't care. Still would prefer an actual artists work and I'd certainly never knowingly pay for something generated by AI.

I don't hate it. I think it's fun as a sort of moment by moment ( I want to see this ) and just generate it and enjoy the wackyness. It does leave a lot to be desired in terms of composition and polish. I also absolutely hate people representing it as their own work. I also really enjoy art produced by people. I think what people produce is still superior in lots of ways. People are often telling a story with their art, and that really comes through. Also I love knowing the amount of thought and effort has gone into a work it makes it that much more impressive. The art people produce is often strongly influenced by art trends, culture, and life experience which we connect to as humans and AI can't produce that because it has no concept of these things. Sure AI can replicate that but it's not the same as the interaction and conversation I have with a piece of art produced by a person that I know must have felt certain ways about their work when producing it.

No, just see it as another medium . Extremely overhated

Tbf tho lotsa popular styles that show in AI art am indifferent towards (even dislike outright . Example : this's somehow even greater assault on the eyes than Alegria illustrations) , but that's bcus it's really hard to create (unique|distinctive) styles with current tech (source : tried developing style for >1 yr (find|combin)ing artist tags in furry models → (genn|tweak)ing ~20-30 training imgs Once satisfied → testing outputs of style LoRA trained with PixAI and result on merge models don't lꝏk like the training data at all . PAINFUL) and not criticism of genAI itself

I like it quite a bit. Le chat mistral does a good job